SimulatedAnnealingLHS

class SimulatedAnnealingLHS(*args)

LHS optimization using simulated annealing.

Performs the optimization of an LHS using simulated annealing algorithm.

Available constructors:

SimulatedAnnealingLHS(lhsDesign)

SimulatedAnnealingLHS(lhsDesign, spaceFilling)

SimulatedAnnealingLHS(lhsDesign, spaceFilling, profile)

SimulatedAnnealingLHS(initialDesign, distribution)

SimulatedAnnealingLHS(initialDesign, distribution, spaceFilling)

SimulatedAnnealingLHS(initialDesign, distribution, spaceFilling, profile)

Parameters:
lhsDesignLHSExperiment

Factory that generate designs

initialDesign2d-array sequence

Initial design to be optimized

distributionDistribution

Distribution of designs

spaceFillingSpaceFilling

Criterion to be optimized Default is SpaceFillingPhiP

profileTemperatureProfile

Temperature profile used by the simulated annealing algorithm Default is GeometricProfile

Notes

With the first constructor, the initial design is generated thanks to lhsDesign. With the second usage (initialDesign), it must be already generated. Starting from this design, a new design is obtained by swapping one random coordinate of two randomly chosen elements; by construction, this design is also an LHS design. If the new design is better than the previous one, it is kept. If it is worse, it may anyway be kept with some probability, which depends on how these designs compare, but also on a temperature profile T which decreases over time. This means that jumping away from local extrema becomes less probable over time.

Examples

>>> import openturns as ot
>>> dimension = 3
>>> size = 100
>>> # Build standard randomized LHS algorithm
>>> distribution = ot.JointDistribution([ot.Uniform(0.0, 1.0)]*dimension)
>>> lhs = ot.LHSExperiment(distribution, size)
>>> lhs.setAlwaysShuffle(True) # randomized
>>> # Defining space fillings
>>> spaceFilling = ot.SpaceFillingC2()
>>> # Geometric profile
>>> geomProfile = ot.GeometricProfile(10.0, 0.95, 2000)
>>> # Simulated Annealing LHS with geometric temperature profile, C2 optimization
>>> optimalLHSAlgorithm = ot.SimulatedAnnealingLHS(lhs, spaceFilling, geomProfile)

Methods

generate()

Generate points according to the type of the experiment.

generateWithRestart(nRestart)

Generate sample with restart.

generateWithWeights()

Generate points and their associated weight according to the type of the experiment.

getClassName()

Accessor to the object's name.

getDistribution()

Accessor to the distribution.

getLHS()

Return the LHS design.

getName()

Accessor to the object's name.

getResult()

Result accessor.

getSize()

Accessor to the size of the generated sample.

getSpaceFilling()

Return the space-filling criterion to be optimized.

hasName()

Test if the object is named.

hasUniformWeights()

Ask whether the experiment has uniform weights.

isRandom()

Accessor to the randomness of quadrature.

setDistribution(distribution)

Accessor to the distribution.

setName(name)

Accessor to the object's name.

setSize(size)

Accessor to the size of the generated sample.

__init__(*args)
generate()

Generate points according to the type of the experiment.

Returns:
sampleSample

Points (\inputReal_i)_{i = 1, ..., \sampleSize} of the design of experiments. The sampling method is defined by the type of the weighted experiment.

Examples

>>> import openturns as ot
>>> ot.RandomGenerator.SetSeed(0)
>>> myExperiment = ot.MonteCarloExperiment(ot.Normal(2), 5)
>>> sample = myExperiment.generate()
>>> print(sample)
    [ X0        X1        ]
0 : [  0.608202 -1.26617  ]
1 : [ -0.438266  1.20548  ]
2 : [ -2.18139   0.350042 ]
3 : [ -0.355007  1.43725  ]
4 : [  0.810668  0.793156 ]
generateWithRestart(nRestart)

Generate sample with restart.

Randomly generate several samples from an initial state and select the one that has the best score.

Parameters:
nRestartint

Number of restarts

Returns:
sampleSample

The best scored sample across restarts

generateWithWeights()

Generate points and their associated weight according to the type of the experiment.

Returns:
sampleSample

The points of the design of experiments. The sampling method is defined by the nature of the experiment.

weightsPoint of size \sampleSize

Weights (w_i)_{i = 1, ..., \sampleSize} associated with the points. By default, all the weights are equal to \frac{1}{\sampleSize}.

Examples

>>> import openturns as ot
>>> ot.RandomGenerator.SetSeed(0)
>>> myExperiment = ot.MonteCarloExperiment(ot.Normal(2), 5)
>>> sample, weights = myExperiment.generateWithWeights()
>>> print(sample)
    [ X0        X1        ]
0 : [  0.608202 -1.26617  ]
1 : [ -0.438266  1.20548  ]
2 : [ -2.18139   0.350042 ]
3 : [ -0.355007  1.43725  ]
4 : [  0.810668  0.793156 ]
>>> print(weights)
[0.2,0.2,0.2,0.2,0.2]
getClassName()

Accessor to the object’s name.

Returns:
class_namestr

The object class name (object.__class__.__name__).

getDistribution()

Accessor to the distribution.

Returns:
distributionDistribution

Distribution of the input random vector.

getLHS()

Return the LHS design.

Returns:
valueLHSExperiment

Result the factory that builds initial design to be optimized

getName()

Accessor to the object’s name.

Returns:
namestr

The name of the object.

getResult()

Result accessor.

Returns:
valueLHSResult

Result of generation that contains the optimal design, some criteria and history

getSize()

Accessor to the size of the generated sample.

Returns:
sizepositive int

Number \sampleSize of points constituting the design of experiments.

getSpaceFilling()

Return the space-filling criterion to be optimized.

Returns:
valueSpaceFilling

Criterion function to be optimized

hasName()

Test if the object is named.

Returns:
hasNamebool

True if the name is not empty.

hasUniformWeights()

Ask whether the experiment has uniform weights.

Returns:
hasUniformWeightsbool

Whether the experiment has uniform weights.

isRandom()

Accessor to the randomness of quadrature.

Parameters:
isRandombool

Is true if the design of experiments is random. Otherwise, the design of experiment is assumed to be deterministic.

setDistribution(distribution)

Accessor to the distribution.

Parameters:
distributionDistribution

Distribution of the input random vector.

setName(name)

Accessor to the object’s name.

Parameters:
namestr

The name of the object.

setSize(size)

Accessor to the size of the generated sample.

Parameters:
sizepositive int

Number \sampleSize of points constituting the design of experiments.

Examples using the class

Kriging: configure the optimization solver

Kriging: configure the optimization solver

Various design of experiments

Various design of experiments

Optimize an LHS design of experiments

Optimize an LHS design of experiments